v1.0 · London · 2026

Simon Rendon Arango

Engineer. Creates software. Uses AI cleverly. Occasionally predicts a football match.

Software and AI engineer. Currently doing AI research and agent enablement at Tessl. On the side I build LFM, a large football model, and EvalLens, an open-source LLM eval framework.

Role
Software · AI engineer
Based
London, UK
Get in touchSee work
or read the constellation →
START HERE · NOTE № 0001Who am I?A first note. Who I am, what this site is, and what software has to look like now that the language models are here.
rendonarango.sh
sra@london
rendonarango.sh · type 'help' or try a chip below.
$
About

Software and AI Engineer

Hi, I’m Simon Rendon Arango, a Software and AI Engineer passionate about building intelligent systems that turn complex data into actionable insights. I hold an MSc in Computing (Software Engineering) from Imperial College London and a BSc in Systems and Computing Engineering from Universidad de los Andes. My professional journey spans startups and fintech, where I’ve designed AI-driven KPI extraction modules, developed scalable backend services, and built user-facing products at companies like Untap, Glamper, and Nequi (Bancolombia). Today I’m at Tessl, working on AI research and agent enablement — building the elements that power autonomous coding services end to end, and the verification systems that check their outputs.

I’m also a curious creator, constantly exploring new technologies and side projects at the intersection of AI, data, and design. I thrive in fast-paced, collaborative environments where ambitious ideas meet rigorous execution — always pushing the boundaries of what’s possible with software and AI.

Selected work

Things I have shipped

Newest first.
  1. Note · № 0002EvalLens
  3. Note · № 0003The constellation
Experience

Where I have worked

Jun 2026 — Present
Tessl

AI Research & Agent Enablement

  • Building the elements that power autonomous coding services end to end — research and engineering across the agent stack
  • Designing verification systems that check and validate agent outputs
  • LLMs
  • Agents
  • Evals
  • TypeScript
  • Python
  • MCP
Nov 2024 — Jun 2026
Untap

Software Engineer

  • Developed and maintained software for private equity investment management using Java, Python, and JavaScript
  • Integrated Google Gemini for intelligent extraction and summarization of business‑critical information
  • Built an LLM‑based system to extract KPIs from financial reports with 90%+ accuracy
  • Implemented automated testing to ensure code quality, reliability, and maintainability across the development lifecycle
Outcomes

90%+ KPI extraction accuracyAutomated testing coverage increasedFaster insight retrieval via Gemini

  • React
  • JavaScript
  • Java
  • Vertex
  • Gemini
  • GCP
Mar 2023 — Sep 2023
Glamper

Software Engineer

  • Developed and maintained the Glamper web application using Vue.js and Nuxt.js, enhancing user experience and functionality
  • Integrated LLMs to provide personalized travel recommendations and improve customer engagement
  • Participated in code reviews and implemented best practices to ensure high-quality code and maintainability
  • Created a serverless backend using AWS Lambda and API Gateway to handle user requests and data processing
Outcomes

LLM‑powered personalizationServerless backend on AWSImproved UX & engagement

  • Vue
  • JavaScript
  • Nuxt
  • Python
  • AWS
Jul 2021 — Jan 2022
Bancolombia

Intern Software Engineer

  • Developed internal AWS–Jira connectors to automate incident tracking, improving response time for infrastructure tickets.
  • Automated several cloud migration workflows, significantly reducing deployment overhead as Bancolombia transitioned to AWS.
Outcomes

Automated incident trackingReduced deployment overhead

  • Python
  • JavaScript
  • AWS
Education

Where I was taught

02 / 02
2018 — 2023
Bogotá, Colombia
Bachelor's Degree

Bachelor of Science in Systems and Computing Engineering

Universidad de los Andes

Core modules

Data StructuresAlgorithm design and analysisSoftware EngineeringMobile and Web DevelopmentSoftware Architecture

Sep 2023 — Sep 2024
London, United Kingdom
Master's Degree

Master of Science in Computing (Software Engineering)

Imperial College London

Core modules

Reinforcement LearningDeep LearningSoftware EngineeringMachine LearningNatural Language Processing

Skills

What I reach for

Core is deepest. Strong is production-ready. Working is honest.
CoreDeepest and most frequently used
14
  • GitDevOps
  • GitHubDevOps
  • JavaScriptFrontend
  • JupyterML · Data
  • LLMsML · Data
  • Next.jsFrontend
  • Node.jsBackend · Cloud
  • NumPyML · Data
  • PandasML · Data
  • PythonBackend · Cloud
  • RAGML · Data
  • ReactFrontend
  • Tailwind CSSFrontend
  • TypeScriptFrontend
Strong
Production-ready, used often
  • AWS
  • CI/CD
  • Data Visualization
  • Deep Learning
  • Design
  • Docker
  • Express
  • Framer Motion
  • GCP
  • GitLab
  • Java
  • MCP
  • Reinforcement Learning
  • Scikit-learn
  • shadcn/ui
  • TensorFlow
Working
Comfortable, actively improving

C++HPCNuxt.jsPyTorchVue

Ask

Ask me anything about my work

A small agent over my CV. It cites what it knows and admits what it does not.

RAG · indexed over CV · citations when available
Product · in progress

Large Football Model

Open LFM

Real-time match probabilities, an embedding explorer, and a Monte Carlo simulation lab. Streaming ingestion, online inference, vector search — a football engine built to read matches as living systems.

live · embeddings · simulationOpen LFM
Open source

EvalLens

Open EvalLens

A systematic evaluation framework for LLM outputs. Reproducible, versioned metrics across model and prompt variants — catches regressions before they ship instead of after.

NOTE · № 0002EvalLensA small open-source library for evaluating structured LLM outputs, with the diff at row level. Vibes are not an evaluation.
reproducible · CI-ready · regressionsOpen EvalLens
This site · editorial

The Constellation

Open the constellation

A career rendered as a field of stars. Four editorial lenses redraw the connections between the same thirteen points — an engineering throughline, a tightening ML loop, a fleet of shipped products, a research compass from Bogotá to London.

NOTE · № 0003The constellationWhy the centrepiece of this site is a hand-drawn map and not a list of jobs in reverse chronological order.
four lenses · thirteen stars · twenty edgesOpen the constellation
From the library

Notes.

Short, illustrated essays. Each one is a single idea, written down so it does not get lost in the work.

All notes →
Contact

Work together

Roles, collaborations, or interesting problems — send a note.